The SMC carbon stars observed were selected from the catalogue of carbon star candidates in the outer regions of the SMC produced by Morgan & Hatzidimitriou (1995). Although a number of carbon star surveys exist for the Magellanic Clouds, they have been, until recently, mostly concentrated on the central regions of the Clouds (see Azzopardi 1993 for a review). Morgan & Hatzidimitriou (1995) have published results of a survey of SMC carbon stars identified on objective-prism photographs taken with the UK Schmidt Telescope. This survey covers a total of , improving by fortyfold on the spatial coverage of previous catalogues and resulting in the discovery of more than one thousand new candidate carbon stars, mostly in the outer regions of the SMC. Most of the SMC carbon stars for which we obtained velocities lie beyond the second to last carbon-star isopleth (of Fig. 6 of Morgan & Hatzidimitriou 1995). Actually we have obtained velocities for 35% of the carbon star candidates identified in these outermost regions.
A small sample of 16 carbon star candidates in (or superimposed on) the outer halo of the LMC, were also observed. These stars were identified on UKST plate YJ13472P which is centred on ESO/SERC Field 32 (4 24, -75 00). The plate is similar to those used by Morgan & Hatzidimitriou (1995) and was scanned as part of a project to identify carbon stars in the LMC.
Table 1a (Tables 1a,b are available electronically) gives the coordinates of the 71 SMC stars observed for which reliable velocities could be determined (see Sect. 3). The projected distances of these stars from the dynamical centre of the SMC (RA: , Dec: (2000), e.g. Brück 1980), are also given in Table 1a. Table 1b gives the coordinates of the LMC stars observed and their projected distances from the centre of the SMC. We used the centre of the SMC as a reference point for these stars as well, in order to be able to compare them readily with the outer Wing/intercloud stars included in the "SMC'' sample.
Figure 1 (click here) shows the spatial distribution of the observed stars. The contours represent the overall distribution of carbon stars in the SMC derived by Morgan & Hatzidimitriou (1995). It can be seen immediately that the majority of the stars observed lie in the outer parts of the SMC.
Figure 1: The spatial distribution of the observed stars, in the SMC (stars MH 1177, MH 1181 and MH 1185 are beyond the margins of the figure). The contours represent the overall distribution of carbon stars in the SMC from Morgan & Hatzidimitriou (1995). Different symbols were used for the different velocity ranges, as indicated
The data were obtained during two observing runs, the first in November 1993 (three nights) and the second from the end of December 1995 to early January 1996 (5 nights), with the Australian National University's 2.3 m telescope at Siding Spring Observatory. We used the Double Beam Spectrograph. In the 1993 observing run, the detectors employed were two thick Loral CCDs, while in the 1996 observing run, the CCDs had been replaced with higher quantum efficiency thinned anti-reflection coated SITe chips. The 1993 spectra covered the following wavelength regions: at a dispersion of 1.1 Å/pixel (corresponding to a radial velocity of 64 km/s/pixel) and at a dispersion of 0.53 Å/pixel (corresponding to a radial velocity of 19 km/s/pixel). The wavelength coverage was larger in the second run, due to the larger size of the CCDs used ( for the blue spectra, and for the red spectra), while the dispersions were the same.
During the first run, the seeing was arsec on the first two nights and 1.5-3 arcsec on the third night. During the second run the weather conditions were poorer (though the seeing was still good, arcsec, in most cases), so a smaller percentage of the time available was usable, hence the small number of new observations, despite the increased quantum efficiency of the CCDs. In the last column of Table 1 we note the observing run during which each spectrum was obtained.
Typical exposure times ranged from 300 to 1200 s for both the blue and the red spectra. The maximum length of a single exposure was 600 s, while multiple exposures were secured for the fainter stars. A series of carbon star radial velocity standards were also observed, in both observing runs: details are given in Table 2 (click here).
Table 2: The carbon stars used as velocity standards. Column 1 gives the names of the stars, Columns 2 and 3 their coordinates (1950), Column 4 the radial velocity adopted for each star (using values from the bibliography, obtained by a search in the SIMBAD database), Column 5 shows the mean observed radial velocity for each star (average from both runs), Column 6 gives the number of observations obtained for each star, and Column 7 denotes the run during which the observations were made
The blue spectral region included the three Swan bands of molecular C and provided immediate confirmation that the programme objects were indeed carbon stars.
The spectra were reduced using the FIGARO package (Shortridge 1990). Standard techniques were applied to correct for the bias offset and pixel-to-pixel response. As most stars had two consecutive observations, it was possible to remove cosmic ray events using the technique described by Croke (1995). For single observations, cosmic ray events were removed using the CLEAN routine of the FIGARO package. When cosmic rays were located near night-sky emission lines, the image was sky-subtracted before attempting to remove the cosmic rays. The spectra were extracted and sky subtracted using the optimal extraction algorithm of Horne (1986). Programme stars and standards were reduced in exactly the same manner.